Estimativa de produtividade de cana-de-açúcar no estado de São Paulo através de um modelo agrometeorológico espectral

Abstract

The influence that the current dynamics of the world economy exerts on agribusiness requires that both private and government monitor closely the behavior of the crops, including large scale. The latest advances in geotechnology and remote sensing expanded substantially the monitoring capacity, allowing also the application of agrometeorological models to estimate and predict crops. Recent studies indicate that there is potential to improve these models aiming operational applications. For this reason, this study objectified estimate the productivity of the sugar cane, through a spectral agrometeorological model, in Barretos and Morro Agudo cities, in São Paulo state, having as input variables the ten-days-interval products of DMP, NDVI S10 derivatives of the VEGETATION-2 sensor, with resolution of 1km, and the product LSA SAF MET ETo, got from the images of the SEVIRI sensor aboard the METEOSAT-9 satellite, and output data of the ECMWF european model, with spatial resolution of 4 and 5 km and temporal of 30 minutes. The selected period were the crop-years of 2009/2012 and 2010/2011. After organizing the database necessary to feed the model, were developed routines in scripts to streamline the data processing and extraction. The model results show that there was a variation of 0.16 t/ha in average agricultural productivity of sugarcane in Barretos during the analysis period, which is confirmed by official data provided by the IBGE showed no interannual variation otherwise the CONAB data showed a variation of 3 t/ha less in the second year/harvest analyzed. Back in Morro Agudo, a decline of about 4 t/ha in average productivity in the analyzed period, behavior was also observed by the data of sugar and alcohol plants in the region, provided by CONAB differing only in that estimated volume was approximately 6 t/ha. The results indicate that the model proposed has potential to be explored, and that more studies must be done to amplify the period of analyze in different regions, and if it’s possible using the field data to calibrate the data.